Premiere crashes (boot.ini issue)
 

Hi guys,

I have an AMD 64 Dual 3800+ and sometime ago I was asking people here how could I take my performance to the max, so somebody suggested making some changes on the boot.ini file. I did the suggested changes and to be honest I didn't notice any improvement...but now from time to time I'm working on premiere and the entire system crashes. The pc goes down like there was a power failure. I'm thinking that by enabling the system to use 3gb instead of 2gb (which is the original command on the original boot.ini file) the system overheats after some time, which makes it crash. Now here comes the tricky part. I had the original boot.ini file saved on my hard disk, but I had some serious trouble and it got deleted...not I'm stuck with the changed boot.ini and I cannot keep working like this.
Here is the changed boot.ini file I have right now:

[boot loader]
timeout=0
default=multi(0)disk(0)rdisk(0)partition(1)\WINDOWS
[operating systems]
multi(0)disk(0)rdisk(0)partition(1)\WINDOWS="Microsoft Windows XP Professional" /noexecute=optin /fastdetect
signature(83741ba5)disk(0)rdisk(0)partition(1)\WIN2="Microsoft Windows XP Professional" /fastdetect
multi(0)disk(0)rdisk(1)partition(2)\WINDOWS="Microsoft Windows XP Professional" /fastdetect /3GB /USERVA=3030

If someone knows how can I manually change these settings back to the original to restore things to peace. I really don't know but I'm thinking the boot.ini file is probably the same for most. Anyway, any help will be appreciated.

In the last line there is a "/3GB" (without quotes of course)

Just delete the "/3GB" and reboot.

Yes, that is my guess too (the "USERVA=3030" too), but this is something I need to be absolutely sure about...otherwise the pc may not even start at all (it's what happened last time).

Yes, sorry, I researched this and of course you are correct. Get rid of

/3GB /USERVA=3030

Mine looks like this:

[boot loader]
timeout=30
default=multi(0)disk(0)rdisk(0)partition(1)\WINDOWS
[operating systems]
multi(0)disk(0)rdisk(0)partition(1)\WINDOWS="Microsoft Windows XP Professional" /noexecute=optin /fastdetect

Mine shows "timeout=0" instead of "timeout=30". Do you think this might be affecting things somehow?

Never mind. I found out that the amount you introduce in front of "timeout" is the amount of seconds you have to select from different operative systems you might have installed on your machine.

Anyway, I deleted the /3GB /USERVA=3030 . I'm going to try to export the project I was trying to on premiere to see if this was the problem (I'll be crossing ALL my fingers!).

Apparently it has nothing to do with the boot.ini changes I've made in the past. The error persists. I've just reinstalled premiere to see what's going on...but I'm almost 100% sure it's not this, because there's a couple of applications where this happens too (dvd burners, anti-virus). I hope it's not some mean virus.

the 3gb switch create so many errors, bugs and crashes for me in premiere it wasn't even funny. Of course reverting it did the trick for me, but with the 3gb switch was a total nightmare.

This makes it sound like you have a hardware problem - e.g., bad memory or a poorly-seated card. If the problem seems random but particularly when the CPU is running near full capacity, overheating could be the issue. Even if the heat output is acceptable, failing hardware can often behave just fine when cool and start flaking out when it gets hot....

I would try removing memory sticks (take one out at a time) and stress-test your system to see if the crashing goes away. Similarly, I'd remove any interface cards that aren't needed, just to troubleshoot. Sometimes just removing and reseating cards can help - especially if the contacts have become dirty or they are loose.

Thanks, John. I just read your post and to be honest I've been busy doing exactly what you say here. Restoring the boot.ini didn't do anything. I DO think it might be something related to overheating. I removed my graphic card and reinstalled it again but no go. I also restored the system to a previous time that I save when everything was working and I reinstalled windows. Things are getting complicated.

I used to have 2 hard drives (300gb each) and now I have 3. Do you think this might be generating enough extra heat to make the pc shut down like a power outage? If so, how can I deal with it? Is there a specific cooling system to deal with this situation?

In the meanwhile I'm going to try removing the memory and putting is back in again.

This sucks, because I just finished editing a short film and I can't even export it.

I removed and inserted all my hard drives and memories, but now I cannot try premiere because when I restored the system it deleted some important stuff. It won't let me delete either. I'm going to make a fresh start and format my main HD.

Btw, I've downloaded Speedfan to see if it's a matter of overheating or not, but I never used a program like this before. I made a print screen of the results:

[img=http://img385.imageshack.us/img385/8417/speedfanresultsgr9.th.jpg]

I think 72C is probably not a good temperature.

Good lord! You're absolutely right, 72C is NOT healthy at all. You should invest in some cooling equipment ASAP before you damage something.

Any suggestions?

Actually, if you look at the picture I've posted you'll see that the place where it puts 72C it says "temp3" (which I have no idea what it is). But the temperature on the hard drives is between 32C and 41C (which appears to be correct). What I really want to know is what the heck is "temp3". All the other temperatures seem to oscilate while I monitor them, except "temp3". I don't know if there's any SPEEDFAN users around here that can give me a hand with this...

My pc came with a tool to monitor and control this. It's called Easy Tune5 and when I go to "pc health" I have the following options:
___________________________________________________________________
FAN LOW LIMIT
CPU-->2000 rpm
System-->1000 rpm
Power-->1000 rpm
___________________________________________________________________
TEMPERATURE TRESHOLD
CPU-->60c
System-->60C
___________________________________________________________________

Does anyone know if this is correct? I am allowed to change them.

My PC actually has a digital thermometer readout on the front panel. I run about 45C.

"temp3" is perhaps just a failsafe setting somewhere in your motherboard's BIOS. I doubt your PC is running that hot, or anywhere near that, based on your other temps. I've also found that a lot of the software monitors do not accurately decode the system's temps from the BIOS-reported sensors... and can read inaccurately. Even the motherboard's own BIOS utility (PC Heath, etc.) often reports inaccurate readings for it sensors. Sometimes a true probe is the only reliable solution for accurate temperatures. The real delta (temp. difference. idle, loaded, etc.) is usually more important to know on your components instead of a certain exact number.

But enough of the techno-blather...

A heatsink connectivity issue or failing CPU fan can lead to intermittent crashes/ reboots, esp. under CPU load like rendering. When in doubt, pull of the CPU heatsink/fan and check for that very thin seal of thermal paste. You can read tomshardware.com or anandtech.com for more info/ insight on these issues.

HTH, and good luck!
Brian Brown

Hi Rafael

I had a similar problem on a "non-editing" PC - the case was too hot to touch!! To check whether it's a heat issue, try running with the cover off and a desk fan blowing air over the innards - it isn't a permanent solution, but it might sort your immediate export problem, and confirm whether heat is indeed the problem!

Be careful running with the case open though!!!

Hope this helps

I found out that my hard disks are running between 32C and 42C and that the "temp3" was simply a reference number (as a matter of fact it never changes). Now I still don't know what is causing this. Can it be that the cpu is too hot? Motherboard? Which is a good software to monitor cpu/motherboard temperature? And what is the "normal" temperature for them?

As of now I'm working without the pc's "carcass". I also checked the fans and they are all working, BUT I do feel the base of motherboard fan is a bit hotter than it should be. And also there's a "metal thing" near the graphic card that feels a bit hot too...as a result this seems to be warming up the graphic card. I don't think I have any choices here. I will have to take my pc to a repair shop...which sucks, because I don't trust anyone with my pc. For almost a decade no matter how hard I always managed to fix everything on my pc...but I have to throw down the towel this time.

I installed 2 demos to try to pinpoint what is going on.
Sensor View, which monitors the following temperatures:
CPU
SYS
AUX
HD0
HD1
HD2
CPU CLOCK

And then BurnInTest, which performs a test using ALL components at the same time.

While I performed this test all the temperatures on Sensor View were normal. They went up a little bit, but nothing out of the ordinary...except AUX, which as raising fast. I actually stopped the test when the AUX temperature was at 62C. The problem is that I have no idea what AUX is, so I still cannot tel exactly where the problem is.

AUX can vary. What motherboard and graphics card do you have?

Can you take a photo of the motherboard and highlight the hot part?

The "metal thing" is probably the heatsink over the chipset (often called the North Bridge). If there's no fan on this, you could rig one up, or blow one in the vicinity of it. Most of the major motherboard vendors will now have a fan over the chipset.

A big 120mm case fan can work wonders on airflow without creating too much high-pitched noise. I have one blowing across my five hard drives and it keeps their temps under control.

HTH,
Brian Brown

Ok, I have uploaded 2 pictures of the places that get hot. The red square is the exact location where the heat is concentrated.

http://img454.imageshack.us/my.php?i...hiccardvq5.jpg

http://img454.imageshack.us/my.php?image=fanqn7.jpg

Yup, that's your heatsink on the chipset. Use a fan per my previous post.

Also, dust looks to be pretty bad on your CPU heatsink. Use a small vacuum (with a grounding strap!) and get rid of it. Get to your power supply and everywhere else, while you're at it. Excessive dust will lead to overheating problems.

BTW, those compressed air canisters just blow the dust around (to get sucked right back up by your PC). That's why I recommend a vac.

HTH,
Brian Brown

Thanks, Brian. Tomorrow is a holiday here in Spain, so everything will be closed, but on tuesday I'm buying a better cooling system for sure. I'll let you guys know when things are back to normal.

If there was any doubt at all it was an overheating issue, now I´m absolutely sure. I turned on a full sized fan at full blast and pointed it right at the pc´s interior, then I tried to export the videos from premiere. It managed to finish exporting for the first time. After it was done I touched the fan I posted the picture of (the one on the motherboard) and it was VERY hot. Tomorrow I have some "cool" shopping to do.